Characterization of cochlear degeneration in the inner ear of the German waltzing guinea pig : a morphological, cellular, and molecular study

(1)

From the Department of Clinical Neuroscience and the Center for Hearing and Communication Research,

Karolinska Institutet, Stockholm, Sweden

CHARACTERIZATION OF

COCHLEAR DEGENERATION IN THE INNER EAR OF THE GERMAN

WALTZING GUINEA PIG: A MORPHOLOGICAL, CELLULAR,

AND MOLECULAR STUDY

Zhe Jin

Stockholm 2006

(2)

All previously published papers were reproduced with permission from the publisher.

Published and printed by Karolinska University Press Box 200, SE-171 77 Stockholm, Sweden

(3)

To my beloved family

(4)

(5)

ABSTRACT

The German waltzing guinea pig is a new strain of animals with yet unknown gene mutation(s) displaying recessively inherited cochleovestibular impairment. The homozygous animals (gw/gw) are deaf already at birth and display typical waltzing behavior throughout life. The heterozygous animals (gw/+) do not suffer from hearing loss and vestibular symptoms. The present thesis is focused on the homozygous German waltzing guinea pig (gw/gw) and its cochlear deficits.

Cochlear histological analysis in the postnatal animals revealed a characteristic cochleosaccular defect: collapse of scala media compartment, atrophy of stria vascularis, degeneration of sensory hair cells, and slow loss of spiral ganglion neurons.

The stria vascularis was atrophic and appeared as a single layer composed of only marginal cells, as further evidenced by morphometric measurement of strial height and width. The abnormally appearing strial intermediate cells (melanocytes) were sparsely scattered in the spiral ligament, whereas strial basal cells were difficult to identify. The degree of sensory hair cell degeneration varied even among animals of the same age.

Morphometric analysis of the spiral ganglion neuron profile density showed a significant reduction in the old (1-2 years of age) animals.

An ensuing morphological study in the cochlea of prenatal embryos showed a progressive reduction of the scala media from embryonic day (E) 35 and its complete absence by E50. The degeneration of hair cells was first observed at E50 and onwards.

The immature stria vascularis failed to transform into a multilayered epithelium but consisted of one layer of underdeveloped/degenerated marginal cells. Strial intermediate cells were sparsely distributed in the spiral ligament and showed signs of degeneration. Strial basal cells were not easily recognized. RT-PCR analysis of the expression of genes regulating strial melanocyte development showed that Pax3 mRNA was remarkably decreased in the cochlear lateral wall but remained intact in the diaphragm muscle and skin tissue.

The loss of the endolymphatic compartment in the gw/gw cochlea suggests a disruption of cochlear fluid and ion homeostasis. The mRNA and protein expression of several key players in cochlear homeostasis were thus investigated in the cochlear lateral wall by semi-quantitative RT-PCR and immunohistochemistry. RT-PCR analysis showed a significant reduction in expression of the strial intermediate cell- specific gene Dct and the tight junction gene Cldn11 in the gw/gw cochlear lateral wall.

Immunohistochemical analysis of the gw/gw cochlea showed loss of the tight junction protein CLDN11 in strial basal cells from E40, loss of the potassium channel subunit KCNJ10 in strial intermediate cells from E50, and loss of the Na-K-Cl cotransporter SLC12A2 in strial marginal cells from E50.

In conclusion, dysfunctional strial cells, in particular intermediate cells, fail to maintain the integrity of the stria vascularis and eradicate the main cochlear K⁺ recycling pathway in the German waltzing guinea pig inner ear, ultimately resulting in the disruption of cochlear homeostasis and cochlear degeneration.

Key words: Cochlea; Deafness; Embryonic development; Inner ear; Kir4.1;

Melanocyte; Pax3, Phenotype; Recessive genes; Recycling; Stria vascularis

(6)

LIST OF PUBLICATIONS

This thesis is based on the following papers, which are referred to in the text by their Roman numerals:

I. Jin Z, Mannström P, Skjönsberg Å, Järlebark L, Ulfendahl M. (2006) Auditory function and cochlear morphology in the German waltzing guinea pig. Hear Res, 219: 74-84.

II. Jin Z, Mannström P, Järlebark L, Ulfendahl M. (2006) Dysplasia of the stria vascularis in the developing inner ear of the German waltzing guinea pig.

Submitted to Cell Tissue Res.

III. Jin Z, Wei D, Järlebark L. (2006) Developmental expression and localization of KCNJ10 K⁺ channels in the guinea pig inner ear. NeuroReport, 17: 475- 479.

IV. Jin Z, Ulfendahl M, Järlebark L. (2006) Disruption of cochlear homeostasis in the developing inner ear of the German waltzing guinea pig: Molecular and cellular studies. Submitted to J Neurosci Res.

(7)

PREFACE

This thesis describes the most recent findings on a long and winding path to unravel the full story behind the inborn deafness of the German waltzing guinea pig. I believe that this thesis will most definitely be read through by the dissertation opponent and thesis committee, and hopefully some of my colleagues (you had better read it!), but it is not very likely to reach a more general audience.

However, for you, and those laymen who may not be familiar with auditory neuroscience and still find their way to these pages, the introduction section will provide some essential information as well as some clues to why these studies were performed and the papers written, and why studies like these should be done in the future. I sincerely hope that all readers will find this a fascinating, albeit seemingly never-ending story... just like I do.

Zhe Jin, Stockholm

(8)

(9)

LIST OF ABBREVIATIONS

+/+ Wild-type animal(s)

ABR Auditory brainstem response

ANOVA Analysis of variance between groups

Ca²⁺ Calcium ion

Cl^- Chloride ion

CSF Cerebrospinal fluid

Dct Dopachrome tautomerase

E Embryonic day

EP Endocochlear potential

gw/+ Heterozygous animal(s) of the German waltzing guinea pig strain gw/gw Homozygous animal(s) of the German waltzing guinea pig strain

HCO3- Bicarbonate ion

K⁺ Potassium ion

KID Keratitis-ichthyosis-deafness

oc Organ of Corti

P Postnatal day

rm Reissner’s membrane

RT-PCR Reverse transcriptase-polymerase chain reaction

SGN Spiral ganglion neuron(s)

SM Scala media

SSH Suppression subtractive hybridization

ST Scala tympani

Stv Stria vascularis

SV Scala vestibuli

TUNEL Terminal transferase dUTP nick end labeling

(12)

(13)

1 INTRODUCTION

The mammalian ear is a unique sensory system to perceive sound in the environment. It is composed of three major portions: the outer, middle and inner ear (Fig. 1). Each portion exerts a specific function in the hearing processes. The outer ear includes the auricle and the external auditory canal, which collects and transfers sound to the middle ear. The middle ear is an air-filled cavity containing the tympanic membrane and three tiny bones known as the ossicles (malleus, incus, and stapes), which serve as a mechanical lever that amplifies the vibration of tympanic membrane into the pressure wave in the inner ear. Deeply embedded inside the temporal bone, the inner ear constitutes the hearing (cochlea) and balance (vestibule and three semicircular canals) organs. These sensory organs of the inner ear are filled with fluids and contain sensory receptor cells – hair cells.

How can we hear? Sound waves are captured by the auricle and travel through the external auditory canal to vibrate the tympanic membrane. The vibrations are conducted to the oval window of the cochlea via the ossicle chain. The subsequent movement of the oval window creates pressure waves in the cochlear fluids and causes the basilar membrane to vibrate. Sensory hair cells sense the mechanical vibration and convert it into electrical signals. These signals are delivered by the auditory nerve up to the brain, where the signals are interpreted as sound.

Figure 1. Anatomy of the human ear consisting of the outer, middle and inner ear. The inner ear contains the sensory organs for hearing (cochlea) and balance (vestibule and semicircular canals). (adapted from http://www.gpc.edu/~jaliff/anasense.htm)

(14)

1.1 MAMMALIAN COCHLEA – HEARING ORGAN 1.1.1 General cochlear anatomy

The mammalian cochlea is a snail-shaped, fluid-filled structure divided by two thin membranes into three parallel compartments: scala tympani, scala vestibuli and scala media (cochlear duct) (Fig. 2). Scala media, containing high [K⁺] endolymph, is separated from scala vestibuli by the Reissner’s membrane and from scala tympani by the basilar membrane. Scala tympani and scala vestibuli, filled with perilymph, are connected by a small opening at the cochlear apex, also known as helicotrema. The mechanoreceptive sensory epithelium, the organ of Corti, resides within scala media on the basilar membrane and extends throughout the length of the cochlea. It contains two types of sensory hair cells: one single row of inner hair cells and three rows of outer hair cells which are separated and surrounded by various supporting cells. On the apical surface of hair cells, actin-based cellular protrusions, stereocilia, are arranged in graded heights to serve as mechanoelectrical transducers. The tectorial membrane, which is a gelatinous structure, is attached primarily to the tips of stereocilia. Extending from primary auditory neurons (spiral ganglion neurons), the nerve fibers mainly contact on the basal surface of hair cells. Spiral ganglion constitutes two types of neurons (type I and type II), which bring electrical signals from hair cells to the brain. Located on the cochlear lateral wall are two non-sensorineural components, the stria vascularis and spiral ligament. The stria vascularis is a multilayered secretory epithelium, while the adjacent spiral ligament is a connective tissue structure containing five subtypes of fibrocytes. Both sensorineural and non-sensorineural components within the cochlea are indispensable for the normal hearing function.

Figure 2. The schematic diagram illustrating a cross-section view of the cochlea. The cochlea is divided into three fluid-filled compartments: the scala vestibuli, media and tympani. The organ of Corti containing sensory hair cells resides within the scala media on the basilar membrane.

The spiral ganglion has two types of primary auditory neurons (type I and II). The stria vascularis and the adjacent spiral ligament are situated along the cochlear lateral wall.

(adapted from Jin and Duan 2006)

(15)

1.1.2 Cellular structure of stria vascularis and its development The stria vascularis is a highly

vascularized epithelium with high secretory and metabolic activity, which is mainly composed of marginal, intermediate and basal cells together with capillary endothelial cells (Fig. 3). The characteristic cellular architecture of the mature stria vascularis appears quite similar across different mammalian species, including mouse (Smith 1957), cat (Hinojosa and Rodriguez-Echandia 1966), guinea pig (Engström et al. 1955), and human (Kimura and Schuknecht 1970a; Kimura and Schuknecht 1970b).

The luminal side of stria vascularis facing the scala media is composed of

ectoderm-derived marginal cells, which possess extensive basolateral membrane infoldings and a rich population of mitochondria. On the other side of stria vascularis, the mesoderm-derived basal cells are spindle shaped, arranged tangentially and form a continuous barrier layer separating the stria vascularis from the adjacent spiral ligament. The intermediate cells are neural crest-derived melanocytes, which can easily be identified by their melanin pigments. The intermediate cells as well as blood capillaries are sandwiched between the marginal and basal cell layers. Between the adjacent marginal cells and basal cells are tight junctions, which insulate the stria vascularis as a unique compartment.

The morphogenesis of the stria vascularis has been extensively described in humans (Chiba and Marcus 2000; Lavigne-Rebillard and Bagger-Sjöbäck 1992) and other mammals including guinea pigs both in vivo (Fernandez and Hinojosa 1974; Kikuchi and Hilding 1966; Thorn and Schinko 1985) and in vitro (Mou et al. 1997; Ågrup et al.

1996). The development of the stria vascularis in different species usually occurs with the successive appearance of the marginal, intermediate and basal cells and follows a base-to-apex gradient along the cochlear spiral, although the total time period of development varies among species. The primordial stria vascularis can be distinguished as a multilayered epithelium composed of undifferentiated cells. A condensation of mesenchymal cells develops adjacent to the epithelial layer and a continuous basal lamina lies in between. The first signs of early differentiation in the stria vascularis are the onset of basal lamina degradation and the presence of identifiable intermediate cells (melanocytes). Marginal cells within the epithelial layer appear more cuboidal in shape, and future basal cells are orientated in parallel to the surface of the developing stria vasularis. When the basal lamina is largely degraded and more intermediate cells are incorporated, the marginal cells develop their cellular projections containing a certain number of mitochondria toward the intermediate and basal cells. The flattened basal cells form a distinct layer separating the stria vascularis from the adjacent spiral ligament. Blood capillaries are also packed in the stria vascularis. As the interdigitation between different strial cells proceeds, the stria vascularis finally develops to its mature

Figure 3. The cellular architecture of the mature stria vascularis.

Figure 3. The cellular architecture of the mature stria vascularis (paper II).

(16)

form. The migration of melanocytes from the neural crest to the stria vascularis is essential for the development and interaction of other strial cells (Cable et al. 1992;

Steel and Barkway 1989).

1.2 FLUIDIC AND IONIC HOMEOSTASIS IN THE COCHLEA 1.2.1 Cochlear fluids and endocochlear potential

The cochlea is a unique organ with respect to the different extracellular fluids filling it.

Cochlear fluids play major roles in cochlear physiology such as transmission of the mechanical stimulus to the hair cells. Three major extracellular fluids have been identified in the cochlea: endolymph, perilymph and intrastrial fluid (Wangemann and Schacht 1996). The chemical composition varies greatly between the cochlear fluids (table 1).

Table 1. Chemical composition and electrical potential of the cochlear fluids (adapted from Wangemann and Schacht 1996).

Endolymph of SM

Perilymph of SV

Perilymph of ST

Intrastrial fluid CSF Plasma

K⁺(mM) 157 6 4.2 2 3.1 5

Na⁺ (mM) 1.3 141 148 85 149 145 Ca²⁺ (mM) 0.023 0.6 1.3 0.8 - 2.6 Cl^- (mM) 132 121 119 55 129 106 HCO3- (mM) 31 18 21 - 19 18 Protein (g/l) 0.38 2.42 1.78 - 0.24 42.38

pH 7.4 7.3 7.3 - 7.3 7.3

Potential (mV) ~80-100 <3 0 ~100 0 0 SM, scala media; SV, scala vestibuli; ST, scala tympani; CSF, cerebrospinal fluid.

The perilymph is a typical extracellular fluid, and its ionic composition is similar but not identical to that of plasma and cerebrospinal fluid. The dominant cation in the perilymph is sodium. Both scala vestibuli and scala tympani are filled with perilymph which communicates at the cochlear apex via the helicotrema. However, the perilymph of scala vestibuli and scala tympani differs in composition and origin (Sterkers et al.

1988): the perilymph of scala vestibuli originates mainly from plasma across a blood- perilymph barrier, whereas the perilymph of scala tympani is partially formed by cerebrospinal fluid (CSF). The movement of perilymph in response to vibration of oval window membrane causes the motion of basilar membrane and in turn hair cell stimulation.

The endolymph is a unique extracellular fluid with unusually high [K⁺] but low [Na⁺] and [Ca²⁺]. Enclosed in scala media, the endolymph has direct contact with several different epithelial cell types including sensory hair cells. It is well known that the endolymph originates from the perilymph through the labyrinthine epithelium, rather than from blood plasma (Konishi et al. 1978; Sterkers et al. 1982). Interestingly, the endolymph in scala media possesses a large positive transepithelial potential with respect to perilymph and plasma, designated as the endocochlear potential (EP). The EP was first recorded by von Békésy (Békésy 1951) and its magnitude is around 80-

(17)

100mV in the mammalian cochleas studied. However, no such high potential has been detected in the vestibular endolymphatic labyrinth or any other mammalian organ. The EP in the cochlea is generally considered to be generated by stria vascularis (Békésy 1951; Tasaki and Spyropoulos 1959) and to serve as a major driving force for sensory transduction. The spiral ligament might also contribute to the generation or maintenance of EP as evidenced by a dramatic reduction of EP in Pou3f4-deficient mice with defect of spiral ligament fibrocytes (Minowa et al. 1999). Several electrophysiological models have been proposed to explain the mechanism underlying the EP generation (Marcus and Thalmann 1980; Offner et al. 1987; Salt et al. 1987).

Nevertheless, the molecular substrate that produces EP has only been identified until very recently. The EP is essentially generated by the potassium channel subunit KCNJ10 (Kir4.1) located in the intermediate cells of stria vascularis (Marcus et al.

2002; Takeuchi et al. 2000).

The intrastrial fluid fills the narrow intrastrial compartment, which is isolated from perilymph in the spiral ligament and endolymph in the scala media by basal and marginal cell layers, respectively. The ionic composition of the intrastrial fluid resembles that of the perilymph containing low [K⁺] but relatively high [Na⁺]. Notably, the intrastrial fluid also maintains a positive voltage potential of ~100 mV relative to the perilymph. It is referred to as intrastrial potential (Salt et al. 1987) and has been assumed as a source of EP.

1.2.2 Cochlear potassium recycling and its molecular correlates

In the cochlea, K⁺ ions provide major charge carriers for hair cell transduction as well as for the EP production. Therefore, cochlear K⁺ homeostasis is essential for maintaining high sensitivity of hair cells and thus for normal hearing function. It has been shown by radioactive tracer experiment that K⁺ ions in the endolymph are derived from perilymph rather than from blood plasma (Konishi et al. 1978). The different epithelial cells lining scala media are coupled by tight junctions, which limit extracellular diffusion of K⁺ ions between endolymph and perilymph. An ensuing question arises as to how K⁺ ions are recycled from perilymph back to endolymph.

Several putative routes for K⁺ recirculation in the cochlea have been proposed so far (Fig. 4) (Kikuchi et al. 2000a; Wangemann 2002a; Wangemann 2002b; Weber et al.

2001). Driven by the high endocochlear potential, K⁺ ions in the endolymph pass through the apical mechanotransduction channels into the sensory hair cells, and then exit the hair cells via K⁺ channels (e.g. Kcnq4, Kcnn2 and Kcnma1) along the their basolateral membranes (Kros 1996). The released K⁺ ions are subsequently taken up by surrounding supporting cells via potassium channels and transporters (e.g. K-Cl cotransporters Kcc3 and Kcc4). With aid of gap junction systems, the K⁺ ions are further transported either medially toward the spiral limbus and back to endolymph (Spicer and Schulte 1998), or laterally toward the spiral ligament (Kikuchi et al. 1995).

Alternatively, K⁺ ions can flow through the perilymph above or below the scala media, or through outer sulcus cells toward the spiral ligament (Chiba and Marcus 2000;

Zidanic and Brownell 1990). The epithelial cell system and the connective tissue cell system are two independent gap junction networks in the cochlea, which are mainly composed of GJB2, GJB3 and GJB6. It has been thought that both gap junction

(18)

networks are participating in cochlear K⁺ recirculation pathway by providing intercellular transport routes (Kikuchi et al. 2000b). Through gap junction networks, K⁺ ions are delivered to the stria vascularis and released from the intermediate cells via the KCNJ10 channel into the intrastrial compartment. From there K⁺ ions were taken up by the basolateral Na-K-Cl cotransporter (Slc12a2) and Na-K-ATPase (Atp1a1/Atp1b2) and secreted back into endolymph by the apical Kcnq1/Kcne1 potassium channel of the marginal cells.

An array of functional proteins including ion channels, cotransporters, ATPases, and intercellular junctions are actively participating the cochlear K⁺ recycling pathway (Fig.

4 and 5). The dysfunction of these key regulators is associated with deafness in humans and mouse mutants.

Kcnq1 and Kcne1 encode the α- and β-subunits of K⁺ channel, respectively. These protein subunits co-assemble to form functional channels and have been detected in the apical membrane of strial marginal cells and vestibular dark cells (Neyroud et al. 1997;

Figure 4. Schematic illustration of cochlear K⁺ recycling pathway and the distribution of key molecular players in the cochlea. Gjb2, gap junction protein, beta 2; Gjb3, gap junction protein, beta 3; Gjb6, gap junction protein, beta 6; Kcc3 (Slc12a6), solute carrier family 12, member 6;

Kcc4 (Slc12a7), solute carrier family 12, member 7; Kcnc1, potassium voltage gated channel, Shaw-related subfamily, member 1; Kcnj16, potassium inwardly-rectifying channel, subfamily J, member 16; Kcnq4, potassium voltage-gated channel, subfamily Q, member 4; Kcnn2, potassium intermediate/small conductance calcium-activated channel, subfamily N, member 2;

Kcnma1, potassium large conductance calcium-activated channel, subfamily M, alpha member 1; Cldn14, claudin 14; Kcnq1, potassium voltage-gated channel, subfamily Q, member 1; Kcne1, potassium voltage-gated channel, Isk-related subfamily, member 1; Slc12a2, solute carrier family 12, member 2; Kcnj10, potassium inwardly-rectifying channel, subfamily J, member 10;

Atp1a1, ATPase, Na+/K+ transporting, alpha 1 polypeptide; Atp1b2, ATPase, Na+/K+

transporting, beta 2 polypeptide; Clcnka, chloride channel Ka; Cldnkb, chloride channel Kb;

Bsnd, Bartter syndrome, infantile, with sensorineural deafness (Barttin); Cldn11, claudin 11.

(19)

Nicolas et al. 2001; Sakagami et al.

1991). Additionally, Kcnq1/Kcne1 K⁺ currents were also recorded in both cell types. Kcnq1/Kcne1 K⁺ channel is by far the sole functional element to secret K⁺ ion across the apical membrane of marginal cells.

The essential role of Kcnq1/Kcne1 K⁺ channel for K⁺ secretion in the cochlea was clearly illustrated in the mouse mutants with a targeted deletion of either Kcnq1 or Kcne1 gene (Lee et al. 2000; Vetter et al.

1996). Genetic mutations affecting either KCNQ1 or KCNE1 are responsible for cardioauditory syndrome (Jervell and Lange-Nielsen syndrome) (Neyroud et al. 1997;

Schulze-Bahr et al. 1997).

Slc12a2 encodes a Na-K-Cl cotransporter which is primarily expressed in the basolateral membrane of strial marginal cells and vestibular dark cells (Crouch et al.

1997). In the cochlea, Slc12a2 is important for effective uptake of K⁺ from the intrastrial compartment. K⁺ secretion from the stria vascularis was abolished in mice with Slc12a2 mutations (Delpire et al. 1999; Dixon et al. 1999). However, no human deafness related to mutations in SLC12A2 gene has been yet identified.

Kcnj10 (Kir4.1), an inward rectifier K⁺ channel subunit is specifically expressed in the strial intermediate cells (Ando and Takeuchi 1999). The time course of its developmental expression was closely correlated to the elevation of EP (Hibino et al.

1997). Loss of EP and partial reduction of cochlear endolymphatic volume and [K⁺] in the Kcnj10-null mice (Marcus et al. 2002), indicate that Kcnj10 channel is essential for EP generation and play a major role in the cochlear K⁺ recycling pathway.

Claudin 11 encoded by Cldn11 gene, a member of the claudin family, is an integral membrane protein and one component of tight junction strands. It is exclusively localized between strial basal cells in the cochlea (Kitajiri et al. 2004b). Claudin 11 co- assembling with other tight junction components connects strial basal cells and forms a continuous barrier separating intrastrial fluid from the perilymph in the spiral ligament.

The important role of Claudin 11 in the maintenance of the electrically isolated intrastrial compartment is evidenced by depression of EP in the Cldn11-null mice (Gow et al. 2004; Kitajiri et al. 2004a).

Connexin 26 (Gjb2), one of major gap junction proteins, is abundantly distributed in both the epithelial cell and connective tissue gap junction systems in the cochlea. Gap junctions, especially connexin 26, provide an intercellular passage for cochlear K⁺ ions and therefore are essential for cochlear function. Mutations in the GJB2 gene account for about 50% autosomal recessive non-syndromic deafness (DNFB1) (Zelante et al.

1997). Two independent transgenic mouse mutants with the mutated Gjb2 gene (Cohen-Salmon et al. 2002; Kudo et al. 2003) display hearing loss, but the EP as well

Figure 5. The cellular localization of key ion transport apparatuses in the stria vascularis.

(20)

as the endolymphatic [K⁺] and volume is not altered. It suggests that Gjb2 might not be essential for EP production and cochlear K⁺ recirculation.

1.2.3 Strial melanocyte and its functional role

Melanocytes have been identified in various sites of the inner ear including the stria vascularis, the cochlear modiolus as well as specific locations of the vestibular end organs (Meyer zum Gottesberge 1988). Generally, otic melanocytes originate from the embryonic neural crest (Hilding and Ginzberg 1977). Their precursors, melanoblasts, migrate from neural crest to the inner ear during early development and ultimately differentiate into melanocytes. The melanoblast is an unpigmented cell, but it can be identified with a marker dopachrome tautomerase (Dct) (Steel et al. 1992). In the stria vascularis, the melanocytes, also known as intermediate cells, contain melanin pigments and extend dendritic processes to interdigitate with adjacent marginal and basal cells. It has been showed that there are two forms of intermediate cells present in the mouse stria vascularis: light and dark intermediate cells (Cable and Steel 1991). The light intermediate cells which are present from birth contain large amount of organelles but very few melanin pigments, whereas the dark intermediate cells which are only observed in the adult stria vascularis, are more heavily pigmented and exhibit pynotic nuclei and contain few organelles. These dark intermediate cells are presumed to be a degenerate form of the light ones.

What is the exact role of strial melanocytes in cochlear function? Although it is still not fully understood, the putative roles of strial melanocytes include 1) normal structural development of stria vascularis. The migration of melanocytes into the stria vascularis is important for the differentiation of marginal cells, the interdigitation between the marginal and basal cells as well as the sustainment of strial capillary network, as evidenced by the findings in the stria vascularis of melanocyte-deficient mutant animals (Hoshino et al. 2000; Steel and Barkway 1989). 2) generation and maintenance of EP.

In the dominant white spotting (W) mutant mice which did not contain strial melanocytes, no EP was generated (Steel et al. 1987). In addition, the presence of strial melanocytes is constantly correlated with a measurable EP (Cable et al. 1994). In mutant mice displaying a progressive degeneration of strial melanocytes (e.g. B^lt light mutant mice), the EP gradually decreased with age (Cable et al. 1993). It indicates that the continued presence of strial melanocytes is required for the maintenance of EP.

Interestingly, the EP is independent of the pigment production ability of strial melanocytes since albino animals still have a normal EP despite of the presence of amelanotic melanocytes in the stria vascularis.

A complex molecular network composed of various genes is participating in the different aspects of strial melanocytes development (proliferation, migration, survival and differentiation) (Fig. 6) (Price and Fisher 2001; Tachibana 2001). The reciprocal interaction between these regulatory genes has been extensively studied. Proto- oncogene Kit and its ligand Kitl are necessary for the survival and/or migration of melanoblasts (Wehrle-Haller 2003). Dominant white spotting (W) and Steel (Sl) mutant mice, which have the mutation in the Kit and Kitl genes respectively, lack strial melanocytes and thus cause hearing impairment (Cable et al. 1994; Cable et al. 1995;

(21)

Schrott et al. 1990; Steel et al. 1992).

Endothelin 3 (Edn3) and its receptor Ednrb, transcription factors Pax3, Sox10, and Mitf have also been considered indispensable for the migration, differentiation, and proliferation of melanoblasts and melanocytes. Among these genes, Mitf plays a central role in the development and function of melanocytes (Price and Fisher 2001; Tachibana 2001).

Mutations in the human EDN3, EDNRB, PAX3, SOX10 and MITF genes result in the hearing loss of several subtypes of Waardenburg (auditory-pigmentary) syndrome (Pingault et al. 1998; Read and Newton 1997).

1.3 ANIMAL MODELS FOR HUMAN HEREDITARY HEARING IMPAIRMENT

Human genetic hearing loss is a highly diverse sensory disorder. It accounts for more than 50% of deafness cases. The loss of hearing could be present as the only clinical manifestation (non-syndromic deafness) or as part of a syndrome (syndromic deafness).

To date, at least 400 different types of genetic deafness have been described. Over the last decade, there has been an astonishing progress in exploring the molecular correlate for each type of genetic deafness. More than 100 different genes and many more loci have been identified to be involved in genetic hearing loss. An up-to-date and comprehensive summary of hereditary hearing loss and the causative gene can be viewed at the Hereditary Hearing Loss Homepage (http://webhost.ua.ac.be/hhh/).

However, the limited access to the human inner ear and availability of human temporal bone specimens, impedes further understanding of human pathology of inherited deafness and the underlying molecular mechanisms. In addition, developmental studies and detailed electrophysiological analysis such as endocochlear potential measurements are not readily carried out in human inner ears. In contrast, animal models do not have these disadvantages and especially those animal mutants closely mimicking human hearing disorders are proving extremely useful to aid in the discovery of human deafness genes. A typical example is that the discovery of mouse gene Myo15a for shaker 2 (sh2) mutants leads to the identification of the orthologous human gene MYO15A and human deafness loci for DFNB3 (Wang et al. 1998). Among the different animal models, the mouse has long been considered as a major choice of model in the field of auditory research. It has several advantages such as a short gestation period of around 3 weeks and the applicability of genetic manipulation technologies. According to studies on inner ear pathology of deaf mouse mutants, they can be mainly classified into several groups (Steel 1995): 1) morphogenetic defects, e.g. fibroblast growth factor 3 (Fgfr3) knockout mice. 2) peripheral neural defects, e.g.

Neurotrophin receptor B (Trkb) knockout mice. 3) neuroepithelial defects, e.g. Shaker 1 (sh1) mice. 4) cochleosaccular defects (abnormal endolymph homeostasis), e.g.

potassium channel gene Kcnq1 knockout mice. 5) membrane matrix defects, e.g. alpha Figure 6. Part of the molecular network regulateing melanocyte development.

Mitf gene (purple) and its protein (red) are master regulators.

(22)

tectorin gene (Tecta) knockout mice. This thesis primarily focused on animal mutants with cochleosaccular defects.

1.3.1 Animal mutants with abnormal endolymph homeostasis

A specific type of human inner ear defect also known as Scheibe’s dysplasia (cochleosaccular defect) was first described in congenitally deaf patients (Scheibe 1892). This is caused by the underdevelopment of the inferior part of the membranous labyrinth which forms the cochlear duct and saccule, although the bony labyrinth is fully developed (Ormerod 1960). Despite the presence of some variability, the typical cochlear pathology in Scheibe’s dysplasia/cochleosaccular defect includes a primary stria vascularis defect, reduction/absence of cochlear duct, and variable loss of sensory hair cells and spiral ganglion neurons, indicative of abnormal endolymph homeostasis (Fig. 7). Scheibe’s dysplasia is the most common form of inner ear aplasia associated with congenital deafness in humans (Paparella and Schachem 1991). It is more frequently observed in temporal bones of patients with syndromic deafness, such as Waardenburg (auditory-pigmentory), Usher (auditory-visual), and Jervell and Lange- Nielsen (cardioauditory) syndromes. The molecular correlates for several types of syndromic deafness have been disclosed recently (Willems 2004). However, cochleosaccular defect is rarely associated with non-syndromic hearing loss.

Cochleosaccular defects have been widely reported in a variety of animal species including cat, dog, rat and mouse (table 2). Although there are both intra- and interspecies differences in cochlear pathology, descent of Reissner’s membrane, cochlear duct reduction and atrophy of the stria vascularis are still early and prominent features.

One group deaf animal mutants (only mice so far as I know) carries mutated or eliminated version of target genes which are known to be essential for cochlear K⁺ recirculation, such as Kcnq1, Kcne1, Slc12a2 and Kcnj10. In these postnatal mouse mutants, the endocochlear potential (EP) was reduced or abolished; the cochlear duct

Figure 7. Typical cochlear pathology of Scheibe’s dysplasia / cochleosaccular defect: primary defect in the stria vascularis (stv), sometimes collapse of Reissner’s membrane (rm), secondary degeneration of sensory hair cells and spiral ganglion neurons (sgn). SM, scala media; oc, organ of Corti. (adapted from Steel 1995)

(23)

was diminished to a variable degree following the collapse of Reissner’s membrane.

However, it is not necessary that all the mouse mutants with targeted disruption of genes involved in cochlear ionic homeostasis display the same cochleosaccular pathology. As an example, no obvious morphological malformation was detected in the cochlea of tight junction gene Cldn11-null mouse despite of the reduced EP (Gow et al.

2004; Kitajiri et al. 2004a).

The other group of deaf animal mutants is associated with melanocyte defects in the stria vascularis as well as in the skin. Therefore, they usually have abnormal coat color, such as white spots or patches. Deaf white cats, Dalmatian dogs, white spotting (W) rats and Steel (Sl) mice are well-known examples of animals with pigment-associated deafness (table 2). Interestingly, the absence of strial melanocytes, the lack of EP and the resulting hearing loss can be highly variable between and within each individual animal: some cochleas have no strial melanocyte and no detectable EP, while in other cochleas strial melanocytes are present only in the pigmented portion of the stria vascularis and a reduced EP can be measured (Cable et al. 1994).

Table 2. Animal mutants with cochleosaccular defects

Mutant name Gene Origin Inheritance References

Kcnq1^-/- mice Kcnq1 T R 1, 2

Kcne1^-/- mice

Punk rocker Kcne1 T, S R 3, 4 Slc12a2^-/- mice

Shaker-with-syndactylism mice Slc12a2 T, S R 5, 6, 7 Kcnj10^-/- mice Kcnj10 T R 8, 9

VGA-9 mice

Microphthalmia mice Mitf T, S R, SD 10, 11,12 JF1 mice

WS4 mice Piebald mice

Ednrb S, T R 13, 14, 15

Steel mice Kitl S SD 16, 17 Dominant spotting mice Kit S SD 18, 19 Lethal spotting mice

Edn3^-/- mice Edn3 S, T R 20, 21

Splotch mice Pax3 S SD 22, 23

White spotting rats Kit S R 24, 25, 26 Deaf white cats Unknown S D 27, 28

Dalmatian dogs Unknown S Unknown 29, 30 Origin: T, transgenic and knockout; S, spontaneous. Inheritance: R, recessive; D, dominant; SD, semidominant. References: 1, (Lee et al. 2000); 2, (Rivas and Francis 2005); 3, (Vetter et al. 1996); 4, (Letts et al. 2000); 5, (Delpire et al. 1999); 6, (Dixon et al. 1999); 7, (Pace et al. 2001); 8, (Marcus et al.

2002) ; 9, (Rozengurt et al. 2003) ; 10, (Tachibana et al. 1992) ; 11, (Hodgkinson et al. 1993); 12, (Motohashi et al. 1994); 13, (Koide et al. 1998); 14, (Matsushima et al. 2002); 15, (Pavan and Tilghman 1994); 16, (Zsebo et al. 1990); 17, (Schrott et al. 1990); 18, (Deol 1970); 19, (Steel et al. 1987); 20, (Baynash et al. 1994); 21, (Mayer and Maltby 1964); 22, (Steel and Smith 1992); 23, (Buckiova and Syka

(24)

2004); 24, (Kitamura et al. 1994); 25, (Hoshino et al. 2000); 26, (Araki et al. 2002); 27, (Heid et al.

1998); 28, (Ryugo et al. 2003); 29, (Mair 1976); 30, (Branis and Burda 1985).

1.3.2 The German waltzing guinea pig

The guinea pig has been long established as an important animal model for research on cochlear physiology due to its frequency sensitivity and its larger cochlear size. Similar to humans, their cochlear structure and function have been fully developed in uterus.

Newborn guinea pigs could hear at birth in contrast to those altricial mammals such as mice, cats and dogs, in which the final stage maturation of cochlea occurs postnatally.

In addition, guinea pig as well as human populations are characterized by genomic heterogeneity, which is unendowed in most of deaf mice mutants occurring in large inbred populations. With the recent completion of low coverage guinea pig genome sequence (http://www.broad.mit.edu/mammals/) and advent of more advanced gene- manipulation techniques, it makes it possible in the near future to use guinea pigs as animal models to explore “deafness gene” for human hereditary hearing loss.

To my knowledge, there have been three known guinea pig strains displaying inherited hearing loss and typical waltzing behavior (table 3). The first strain of waltzing guinea pigs known as the Kansas strain was described already by Ibsen and Risty in 1929 as

“Two related individuals with a tendency to whirl, or waltz, similar to that known in Japanese waltzing mice” (Ibsen 1929). The Kansas strain displayed an autosomal recessively inherited hearing loss already from birth. Cochlear morphology in the homozygous animals showed a normal position of Reissner’s membrane but variable degeneration of sensory and neural structures (Lurie 1939; Lurie 1941). Another strain of waltzing guinea pigs, the NIH stain, has been extensively described by Ernston (Ernstson 1970; Ernstson 1971a; Ernstson 1971b; Ernstson 1972; Ernstson et al. 1969).

The NIH strain followed an autosomal dominant inheritance with a recessive lethal effect (Ernstson 1970). The homozygotes displayed progressive hearing loss becoming more severe 14 days after birth (Canlon et al. 1993; Ernstson 1972). Loss of sensory hair cells was manifested at 2-3 months of age in the homozygotes, followed by degeneration of spiral ganglion neurons, although there was no detectable atrophy of stria vascularis. Stereocilia fusion and cuticular plate protrusion were major pathologies in both cochlear and vestibular hair cells. In addition, unique actin filament rods existed in type I vestibular hair cells of homozygous animals as revealed by transmission electron microscopy (TEM) observations (Sobin and Weraall 1983). Although the auditory function and inner ear pathology have been extensively studied, the causative genetic substrate has not yet been identified for either of the two strains of waltzing guinea pigs.

Table 3. Comparison of different strains of waltzing guinea pigs.

Inner ear pathology Strain Name Inheritance mode Mutated gene Availability

Cochlea Vestibule

Kansas AR Unknown No Yes No

NIH AD, RL Unknown Yes Yes Yes

German AR Unknown Yes Yes Yes?

AR, autosomal recessive, AD, autosomal dominant; RL, recessive lethal.

(25)

The third and most recently described strain of waltzing guinea pigs has also arisen spontaneously. It was originally discovered in an animal facility in Germany in 1996, thus named the German waltzing guinea pig. Two normal-behaving guinea pigs had a litter of four offspring, of which two animals (one male and one female) displayed a typical circling/waltzing behavior. These two waltzing guinea pigs were subsequently interbred and produced two litters of progenies, each comprised of two waltzing animals. All six waltzing guinea pigs were then transferred to Karolinska Institutet (Stockholm, Sweden). Systematical breeding of these waltzing animals together with normal guinea pigs from an outbred Swedish strain, revealed the autosomal recessive mode of inheritance (Skjönsberg et al. 2005; Ernstson, unpublished observation).

However, the underlying genetic substrate is still unknown. Both homozygotes (gw/gw) and heterozygotes (gw/+) of the German waltzing guinea pigs are viable and fertile with normal body size and lifespan. They have pigmented eyes as well as normal coat color varying from yellowish to brown despite occasional white spots.

Heterozygotes (gw/+) of the German waltzing guinea pig can not be distinguished from wild-type guinea pigs in external appearance and cochlear morphology. Hearing threshold and endocochlear potential (EP) have been shown within a normal range in the heterozygous animals, which display normal behavior and vestibular reaction.

Interestingly and paradoxically, they exhibit resistance to noise exposure, while more susceptible to ototoxic drugs (Halsey et al. 2005; Skjönsberg et al. 2005). The heterozygous German waltzing guinea pigs represent an intriguing animal group for investigating how exogenous (i.e. environmental) auditory stress factors reciprocally act with endogenous (i.e. genetic) elements.

Homozygotes (gw/gw) of the German waltzing guinea pig (Fig. 8) are readily recognized from their littermates at birth by their distinct circling behavior and head- tossing movement. The lack of Preyer reflex and auditory brainstem response (ABR) in neonatal gw/gw animals suggested they were deaf already (Skjönsberg et al. 2005).

Preliminary histology study showed that scala media was diminished in the prenatal gw/gw animal. This finding indicates that cochlear fluid and/or ion homeostasis is severely disrupted in the inner ear of the gw/gw animal. In addition, the vestibular endolymphatic compartment was similarly collapsed. The work in this study focused on the cochlear portion of the homozygous German waltzing guinea pig (gw/gw). In papers I and II, the time course of cochlear degeneration in the postnatal and prenatal gw/gw animal was described in detail. The cellular and molecular basis for disruption of cochlear homeostasis in the developing gw/gw animal was presented in paper III and IV.

Figure 8. Young (postnatal day 21) homozygous German waltzing guinea pig (gw/gw) with yellow- brown coat color and pigmented eyes.

(26)

2 AIMS OF THE STUDY

The long-term goal of this project was to identify the molecular and genetic substrate for inner ear degeneration in the German waltzing guinea pig (gw/gw). The specific aims of the studies included in the thesis were:

− To characterize the cochlear pathology in the postnatal homozygous animals.

− To characterize the time course of cochlear degeneration and explore the underlying mechanism in the prenatal homozygous embryos/fetuses.

− To investigate the cellular and molecular basis of the disruption of cochlear homeostasis in the developing inner ear of the German waltzing guinea pig.

(27)

3 MATERIALS AND METHODS

3.1 EXPERIMENTAL ANIMALS AND TISSUE PREPARATIONS 3.1.1 Experimental animals

Guinea pigs used in the present study originated from two different strains (table 4).

Homozygous (gw/gw) and heterozygous (gw/+) guinea pigs were derived from the German waltzing guinea pig strain, which are not commercially available and bred only at Karolinska Institutet. Heterozygous (gw/+) animals were identified by controlled breeding (Skjönsberg et al. 2005). Wild-type (+/+) guinea pigs used as control and for breeding purpose were derived from a commercially available guinea pig strain (“Sahlin strain”; Bio Jet Service, Uppsala, Sweden). Healthy and pigmented guinea pigs were selected for breeding to produce a sufficient number of embryos/fetuses.

Table 4. Different age of guinea pigs (embryos) used in the present study

Paper number +/+ gw/gw gw/+

I Postnatal animals from newborn up to 2 years of age

II E25, E30, E40, E45, E50, E60, adult Adult III E30, E35, E40, E45, E50, E60, P0, P60 − − IV E30, E35, E40, E45, E50, E60, P0, adult P0, adult +/+, wild-type; gw/gw, homozygote; gw/+, heterozygote; E, embryonic day; P, postnatal day.

All animal experiments and procedures were approved by Swedish Animal Care and Use Committee (approvals N10/01, N11/01, N464/03, and N465/03). Animals were housed in an animal facility with free access to food and water under 12:12 h light dark cycle. As guinea pigs cannot manufacture their own vitamin C, ascorbic acid (vitamin C) was supplemented in the daily drinking water.

3.1.2 Tissue preparations

Postnatal guinea pigs were deeply anesthetized with an overdose sodium pentobarbital.

For histological and immunohistochemical studies, the animals were transcardially perfused with 0.9% saline followed by ice-cold 4% formaldehyde fixative freshly prepared from paraformaldehyde. After decapitation, the temporal bones were removed and the cochleas were dissected out. Some cochleas were immersed in RNAlater^TM RNA stabilization reagent (Qiagen) to avoid the potential RNA degradation, the cochlear bony capsule was stripped off and the cochlear lateral wall/stria vascularis was carefully detached from cochlear sensorineural structures (Fig. 9). Either the whole cochlea or the dissected cochlear lateral wall was further processed for total RNA isolation. The other cochleas were locally perfused with 4% paraformaldehyde fixative through the round window and a small hole at the apical turn, and then immersed in the same fixative (for light microscopy and immunohistochemistry) or 3% glutaraldehyde (for transmission electron microscopy) overnight at 4ºC. To prepare for cochlear cryosection and histological analysis, the cochleas were decalcified with 0.1 M EDTA made in 0.1 M phosphate buffer (PB) until the bony capsules were soft enough for the subsequent processing.

(28)

Timed pregnant guinea pigs were sacrificed and the abdomen cavity was opened to expose the uterus. The embryos/fetuses were immediately taken out and sacrificed without transcardial perfusion. After dissected out from the temporal bones, the cochleas were processed following the same procedure as described above in the postnatal animals.

The tissue preparation procedure was illustrated in the following flow chart.

3.2 COCHLEAR HISTOLOGY ANALYSIS

3.2.1 Whole-mount and flat-mount morphology

The cochleas from neonatal guinea pigs were dehydrated with a graded methanol series (25%, 50%, 75%, and 100% methanol), and then cleared in a 1:2 mixture of benzyl alcohol: benzyl benzoate (Sigma). Cleared cochleas were examined and photographed under a dissecting microscope (Leica) equipped with a digital camera (JVC). The dissected cochlear lateral wall tissue strips were flat-mounted with the marginal cell side up in phosphate-buffered saline (PBS)/glycerol (1:1) on 8-well slides (Erie Scientific), and examined under a light microscope (Zeiss).

Total RNA islation (whole cochlea, lateral wall), RT-PCR

Paper II-IV

Decalcification Dissection

Dissection

Sacrifice, transcardial fixative perfusion (postnatal animals), local fixative perfusion Sacrifice

Guinea pigs/embryos

Fresh cochleas Fixed cochleas

Immunohistochemistry (whole-mount cochlea, flat-mount lateral wall)

Paper III, IV

Light microscopy

Transmission electron microscopy Immunohistochemistry

(cryosection)

TUNEL assay (cryosection) Paper I-IV

Figure 9. Schematic illustration of dissection of cochlear lateral wall tissue from the guinea pig inner ear.

(29)

3.2.2 Light microscopy

The cochleas were washed in 0.1 M PB several times, dehydrated in a graded ethanol series and embedded in JB4 resin (Polyscience). The cochleas were sectioned at 4 or 5- µm thickness with a rotary microtome (Microm HM355S) along the midmodiolar plane. Every third section was mounted on glass slides, stained with 0.1% toluidine blue and examined under a light microscope (Zeiss).

3.2.3 Transmission electron microscopy

The cochleas were post-fixed in 1% osmium tetroxide, dehydrated in a graded series of ethanol, and embedded in plastic Agar 100 resin (Agar Scientific). After polymerization, the third cochlear turn was dissected out and re-embedded on a blank block of Agar 100 for sectioning. Sections at 1-µm thickness were cut on an ultratome (LKB Cryo Nova) and stained with toluidine blue in order to select angles and regions of interest. Ultra-thin sections were then collected, mounted on formvar-covered copper grids, and stained with uranyl acetate and lead citrate. The sections were examined and photographed in a transmission electron microscope (JEOL 1230).

3.3 COCHLEAR MORPHOMETRIC ANALYSIS

All morphometric analyses were performed on the light micrograph images using SigmaScan Pro 4 Image Analysis software.

3.3.1 Cochlear dimension measurements

Different cochlear parameters were measured on one midmodiolar JB4 section from each group of young (5-9 weeks of age) wild-type (+/+, n=6) and homozygous (gw/gw, n=6) animals. The following cochlear parameters were measured inside the bony capsule (Fig. 10): #1, total height of the cochlea from the apical to basal turn; #2, cross- sectional area of the cochlea; #3, the width of the cochlea at the second turn between left and right lateral walls; #4, cross-sectional area of the fluid compartments at the second turn. The parameters for stria vascularis and spiral ligament were measured at the second cochlear turn: #5, the height of stria vascularis; #6, the width of stria vascularis at its mid-height; #7, the height of spiral ligament; #8, the width of spiral ligament at its mid-height; #9, the width of spiral ligament between stria vascularis at its mid-height and the lateral wall; #10, cross-sectional area of the spiral ligament.

Figure 10. Schematic illustration showing the measurements of different cochlear

parameters (paper I).

(30)

3.3.2 Spiral ganglion neuron profile density

To compare the size of spiral ganglion neuron (SGN) population in different animal groups, SGN profile density was estimated within Rosenthal’s canal. The number of SGN (both type I and II) with a clear nuclear profile and the cross-sectional area of Rosenthal’s canal were measured at two locations separated by a half turn on both sides in the first, second and third cochlear turn on six consecutive cochlear midmodiolar sections. The SGN profile density within Rosenthal’s canal was represented as SGN number divided by the cross-sectional area of Rosenthal’s canal. SGN profile densities from five groups of animals were compared: young adult (5-9 weeks of age) +/+, n=9;

young adult gw/+, n=6; young adult gw/gw, n=6; old +/+ (1-2 years of age), n=2; old gw/gw, n=7.

3.4 MOLECULAR BIOLOGY TECHNIQUES 3.4.1 Total RNA isolation

Whole cochleas or micro-dissected cochlear lateral wall/stria vascularis tissues were immediately homogenized with a plastic pestle in TRIzol solution (Invitrogen) or RLT lysis buffer contained in RNeasy Micro Kit (Qiagen). The tissue lysates were further passed through a 20-gauge needle attached to a sterile syringe several times until a homogenous lysate was achieved. Total RNA isolation was subsequently performed following TRIzol or RNeasy Micro Kit protocols. The isolated total RNA was cleaned up with RNeasy MinElute spin column (Qiagen) and subjected to on-column DNase I digestion (Qiagen) to minimize potential DNA contamination. The concentration and integrity of total RNA was determined using the RNA 6000 Nano assay on an Agilent 2000 Bioanalyzer (Agilent Tech). Only RNA sample with a 28S/18S rRNA ratio more than 1.0 was used for RT-PCR experiment (Fig. 11).

3.4.2 Semi-quantitative RT-PCR

Due to its convenience and high reproducibility, single step RT-PCR was used in the present study. RT-PCR was performed using SuperScript ^TM One-Step RT-PCR with Platinum® Taq System (Invitrogen). Guinea pig gene-specific primer pairs were either retrieved from published papers or designed from highly conserved region of human/mouse/rat sequences with online Primer 3 primer design tool

Figure 11. Representative electropherogram and gel- like image showing total RNA preparation with high quality (28S/18S rRNA ratio≥1).

(31)

(http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). Primer sequences are listed in table 5.

Table 5. PCR primer sequences

Gene Primer sequence (5’-3’) f-forward primer, r- reverse primer

Product size (bp)

Annealing temperature (ºC)

Pax3 f-CGTGCCGTCAGTGAGTTCCA

r-CGCTTTCCTCTGCCTCCTTC 101 60

Sox10 f- AGGTGCTCAGCGGCTACGA

r- TTGGGCGGCAGGTAYTGGTC 679 55

Edn3 f- CAGGATTCGTGCCTTGCTC

r- CCGTCTGTTCGGGAGTGTT 324 50

Kitl f- GATCTGCGGGAATCCTGTGA

r-CGGCGACATAGTTGAGGGTTA 98 50

Kit f- CCCACCCTGGTCATTACAGA

r- CCGICCITGAGTRAGGAGGA 420 56

Mitf f- CCCAACGGCAGCAGGTAAA

r- CAGGACGCTCGTGAATGTG 259 55

Dct f- GATTAGTCGGAACTCRAGATT

r- CATTAGTCACYGGWGGGAAG 509 49

Cldn11 f- GACCACCTCCACCAATGACT

r- CCCGCAGTGTAGTAGAAACG 516 55

Kcnj10 f- CCTCATTGGCTGCCAGGTGACA

r- TGCCTTCCTTTTCAGCTTGCTC 487 56

Kcnq1 f- GGATGGAGATTGTCCTGGTG

r- CCTGGCGATGGATGAAGA 309 62

Atp1a1 f- AGCGATTCTTTGTTTCTTGG

r- CACCAGTGAGCGAGGAGTTA 341 50

Atp1b2 f- TATCCTCCTCTTCTACCTCGT

r- GGGCTTGGATAGAGTCGTT 254 50

Slc12a2 f- GGAATGGAGTGGGAAGCA

r- CTTTGGGTATGGCTGACTGA 283 53

Gjb2 f- TCCCCATCTCBCACATCCGGC

r-AAGATGACCCGGAAGAAGATRCTG 232 58

Gjb6 f- GAAGCAGCCTTTATGTATGTGT

r- AGCAGCAGGTAGCACAACTC 206 50

Gjb3 f- ACGAGCAAAAAGACTTTGACT

r- ATCCTGTGGAAGATGAGGTAG 505 53

Gapdh f- GCCAACATCAAGTGGGGTGATG

r- GTCTTCTGGGTGGCAGTGATG 310 60

The contents of a 25-µl RT-PCR reaction were: 12.5 µl of 2× reaction mix [containing 0.4 mM of each deoxynucleotide triphosphate (dNTP), 2.4 mM MgSO4], 50 ng of total RNA, 0.5 µl each of forward and reverse primers (10 µM), and 0.5 µl of RT/Platinum®

Taq enzyme mix. The optimal annealing temperature and PCR cycle number of each primer pair were determined with a gradient thermocycler (PTC-200, MJ research). The following thermal cycling conditions were used: reverse transcription at 45-55ºC for 30 minutes, pre-denaturation at 94ºC for 2 minutes, followed by optimal cycle of denaturation at 94ºC for 15 s, annealing at 49-62ºC for 30 s, extension at 72ºC for 1 minute, and a final extension at 72ºC for 10 minutes. Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) was run as a positive control for each RNA template. A negative control omitting RNA template from reaction was carried out to detect any DNA contamination in the reaction. Absence of genomic DNA in the RNA template was further verified by omitting the RT/Platinum® Taq mix and substituting with 2 units of Platinum® Taq DNA polymerase (Invitrogen). The RT-PCR condition has

(32)

been optimized to yield PCR products within a linear amplification range in order to compare the relative gene expression. Triplicate RT-PCR amplifications were performed from each RNA template. PCR products were electrophoresed on 2%

agarose gels containing ethidium bromide (0.5 µg/ml), and visualized under UV light.

PCR products were directly sequenced after gel extraction with QIAquick Gel Extraction Kit (Qiagen), using DYEnamic ET terminator cycle sequencing kit (AmershamPharmacia Biotech), and an ABI100 model 377 sequencer (KIseq, the DNA sequencing core at Karolinska Institutet).

3.5 IMMUNOHISTOCHEMISTRY

To prepare cryosections, decalcified cochleas were washed with 0.1 M PBS several times and immersed in 30% sucrose overnight for cryprotection. The cryoprotected cochleas were embedded in O.C.T. compound (Sakura Tissue-Tek), snap frozen in dry ice with isopentan. Cochlear cryosections were cut at 12-µm thickness in a cryostat (Leica) and mounted onto poly-D-lysin-coated SuperFrost glass slides (Erie Scientific).

Selected cryosections were re-hydrated in 0.1 M PBS before the immunostaining, whereas whole-mount cochlea and flat-mount cochlear lateral wall/stria vascularis were directly processed for immunohistochemical labeling.

For immunoperoxidase staining, the tissues were pretreated with 0.3% H2O2 for 10 minutes to inactivate the endogenous peroxidase activity. All tissues were incubated with a blocking solution (containing 3-10% normal goat serum, 0.1-0.3% Triton X-100 in 0.1 M PBS) for 1 h at room temperature, followed by incubation with primary antibodies (listed in table 6) at 4ºC overnight. Two different visualization approaches were used: (1) fluorophore-conjugated secondary antibodies (table 6) for 1 h at room temperature; (2) Avidin: Biotinylated enzyme complex (ABC) approach using biontinylated secondary antibodies (table 6), avidin-biotin-peroxidase complex, and peroxidase substrate (e.g. VIP from Vector Lab). In order to evaluate the antibody specificity, negative controls reactions were performed in parallel, by pre-absorption of the primary antibody with excess control antigen peptide, or by omission of primary or secondary secondary antibodies. The immunolabelling reactions were observed and documented under a stereomicroscope (Leica), a light microscope (Zeiss), a fluorescence microscope (Zeiss), or a confocal laser scanning microscope system (Zeiss LSM 510).

(33)

Table 6. Primary and secondary antibodies used for immunohistochemistry.

Antibody Mono/poly clonal

Host Concentration Producer Paper

Primary

Anti-myosin VIIa Polyclonal Rabbit 1:1000 Dr. Tama

Hasson, USA II Anti-NF-L Monoclonal Mouse 1:200 Santa Cruz II

Anti-KCNJ10 Polyclonal Rabbit 1:100/400 Alomone III, IV Anti-TUJ1 Monoclonal Mouse 1:250 Covance III Anti-KCNQ1 Polyclonal Rabbit 1:800 Chemicon IV Anti-connexin 26 Polyclonal Rabbit 1:250 Dr. David

Kelsell, UK

IV

Anti-SLC12A2 Monoclonal Mouse 1:200 Iowa hybridoma bank

IV

Anti-claudin 11 Monoclonal Mouse 1:1 Dr. Alexander Gow, USA

IV

Secondary Cy3-anti-rabbit

IgG Polyclonal Goat 1:2000 Jackson

ImmunnoLab II-IV FITC-anti-mouse

IgG Polyclonal Goat 1:400 Jackson

ImmunnoLab II-IV Alexa 594-anti-

mouse IgG2A Polyclonal Goat 1:600 MolecularProbes IV Alexa 488-anti-

mouse IgG Polyclonal Goat 1:500 MolecularProbes IV Biotinylated anti-

rabbit IgG Polyclonal Goat 1:1000 Vector Lab III Biotinylated anti-

mouse IgG Polyclonal Goat 1:1000 Vector Lab IV

3.6 APOPTOSIS DETECTION TUNEL ASSAY

Cochlear cryosections were prepared for terminal transferase dUTP nick end labeling (TUNEL) assay. The TUNEL assay is a common method to detect apoptosis-induced DNA fragmentation. In the present study, it was performed using ApopTag Plus peroxidase in situ apoptosis detection kit (Chemicon) following the manufacturer’s instruction. In brief, sections were post-fixed in ethanol: acetic acid (2:1) for 5 minutes at -20ºC. Endogenous peroxidase was then quenched in 3% hydrogen peroxide in PBS for 5 minutes at room temperature (RT). After incubation with terminal deoxynucleotidyl transferase (TdT) enzymein a humidified chamber for 1 h at 37°C followed by anti–digoxigeninperoxidase conjugate for 30 minutes at RT, sections were developed with diaminobenzidine (DAB) substrate for 3 minutes at RT and counterstained in methyl green solution (Vector Lab) for 1 minute at 60ºC. Sections were examined under a light microscope (Zeiss). Substitution of TdT enzyme with distilled water was used as negative control. Slides containing normal female rodent mammary gland tissue (1-2% of the total number of cells on the slide are apoptotic) served as positive controls.

(34)

3.7 BIOTIN TRACER PEMEABILITY ASSAY

Guinea pigs (+/+, gw/+ and gw/gw) were anesthetized with an overdose pentobarbital and transcardially perfused with 0.9% saline to flush out blood cells. The cocheas were dissected out and both the round and oval windows were opened in 0.1 M PBS containing 1 mM CaCl2. About 500 µl 10 mg/ml EZ-Link^TM Sulfo-NHS-LC-Biotin (Pierce Chemical) were perilymphatically perfused into the cochleas for 10 minutes followed by washing 5 times with0.1 M PBS containing 1 mM CaCl2. The cochleas were fixed with 10% TCA (Sigma) for 2 hours and further processed for cryosection. In order to visualize the distribution of the biotin tracer, the cryosections were incubated with streptavidin/fluorescein-isothiocyanate (FITC) (1:500, R&D) for 15 minutes and washed several times with 0.1 M PBS. Slides were examined under a fluorescence microscope (Zeiss).

3.8 STATISTICAL ANALYSIS

One-way ANOVA was used for comparison of spiral ganglion cell profile density as well as cochlear parameter measurements. The level of significant statistical difference was set at p<0.001. Statistical analysis was performed with the Origin 6.0 program.

Characterization of cochlear degeneration in the inner ear of the German waltzing guinea pig : a morphological, cellular, and molecular study

From the Department of Clinical Neuroscience and the Center for Hearing and Communication Research,

Karolinska Institutet, Stockholm, Sweden